Phenotype and functional changes of Vγ9/Vδ2 T lymphocytes in Behçet's disease and the effect of infliximab on Vγ9/Vδ2 T cell expansion, activation and cytotoxicity

Thirteen patients with BD (nine males and four females, mean age 42 ± 24 years), classified according to the International Study Group for Behçet's Disease [8] were studied. The activity of BD was assessed by collecting clinical symptoms defined according to the BDCAF score [9, 10] that includes the presence of several manifestations of the disease, by the uveitis scoring system and by the visual activity measurement [11]. At the time of sampling, disease was active in six patients and inactive in seven. In five of the active patients blood for serum and lymphocyte studies was obtained before and after the anti-TNF-α (Infliximab) therapy. All patients were using colchicine (n = 13), and/or low dose corticosteroids (n = 8). Ten healthy volunteers (age range 21 to 47, mean 30 ± 8 years) were enrolled as controls. None of patients or controls were HIV, CMV, EBV infected. Human studies committee approval and individual informed consent from each patient were obtained.

Peripheral blood mononuclear cells (PBMC) were obtained from each individual by separating heparinized venous blood on Ficoll (Euroclone, Wetherby, Yorkshire, UK). The cells were washed in RPMI-1640 medium (Euroclone), and cultured in 24-well plates (Costar, Cambridge, MA, USA) at a concentration of 5 × 10⁵ cells/ml in RPMI-1640 supplemented with 10% foetal calf serum (Euroclone) 20 mM Hepes (Euroclone), 2 mM L-glutamine (Euroclone) and 100 U/ml penicillin/streptomycin (Sigma, St Louis, MO, USA) at 37°C at 0,5% CO2. For the expansion of Vγ9/Vδ2 T cells, PBMCs were cultured for 10 days in medium alone or in the presence of 0,5 mM Dimethylallyl pyrophosphate (DMAPP, Sigma, St Louis). After 72 hours, cultures were supplemented with a 0,5 ml medium containing 40 U/ml recombinant human IL-2 (Genzyme, Cambridge, MA, USA). Every 72 h, 0.5 ml medium was replaced with a 0.5 ml fresh medium containing IL-2. After 10 days, cells were washed three times in medium, and expansion of Vγ9/Vδ2 T cells was assessed using a FACScan flow cytometer (Becton Dickinson, Mountain View, CA, USA) by using forward scatter/side scatter gating to select the lymphocyte population for analysis. The Vγ9/Vδ2 T cell expansion factor (EF) was then calculated as described above [6].

Monoclonal antibodies (MOAbs) specific for human surface antigens anti-T cell receptor (TCR) Vδ2 fluorescein isothiocyanate-labelled (FITC; PharMigen, San Diego, CA, USA), anti-CD27 phycoerythrin-labelled (PE, PharMigen), anti-CD45RO quantum red (QR, Sigma), anti-TNF-RII PE (R&D System, Minneapolis, MN, USA) were used. For the evaluation of intracytoplasmic content of perforin (Pf) and IFN-γ, 3 × 105 cells were stained with anti- Vδ2 TCR FITC and, after washing, fixed with 4% paraformaldehyde (Sigma) for 30 minutes at 4°C. After two washes with permeabilization buffer (saponine containing) the cells were incubated at 4°C for 45 minutes with anti perforin Pe antibody (Ancell Corporation, Bayport, MN, USA) or with anti-IFN-γ PE (Euroclone). After washing, the cells were suspended in PBS with 1% foetal calf serum and data were acquired on a FACScan instrument and analyzed using WinMDI version 2.8 software.

The number of Pf and TNF-RII molecules (MESF; molecular equivalents of soluble fluorochrome) was calculated by fluorescence-activated cell sorter analysis of cells stained with saturating amounts of PE labelled anti-TNF-RII and anti-Pf mAb of known PE/protein ratio and comparing the staining with a standard curve of microbeads labelled with defined numbers of PE molecules (Quantum Fluorescence Kit, Sigma). The analysis was done using Quickal Program for MESF Units for Windows.

GrA activity was tested using the synthetic substrate N-α-benzyloxycarbonyl-L-lysine thiobenzyl ester (z-lys-SBzl; Sigma) as previously described [7]. Briefly, 20 μl of supernatants (obtained at the 10^th day of Vγ9/Vδ2 culture) or diluited (1:100 in PBS) sera were coincubated with 35 μl 1 mM z-lys-SBzl and 35 μl 1 mM 5,5-dithio-bis-(2nitrobenzoic acid) (Sigma). After incubation at 37°C for 2 hours and 30 minutes respectively, the absorbance at 405 nm was determined. Esterolytic activity of GrA was reported as mOD units.

In order to examine the effects on Vγ9/Vδ2 expansion, TNF-RII expression, perforin and IFN-γ content, and supernatant levels of GrA, Infliximab (Remicade; Centocor Inc., Malvern, PA, USA; Schering Plough SpA, Segrate (Mi), Italy) was added in the medium at a final concentration of 10, 50 (for 3 days) and 100 μg/ml (for 3 and 10 days).

Five patients with active disease were treated with Infliximab, 5 mg/Kg, by a two hour infusion, at weeks 0, 2, 4 and the patients observed for a further two hours without adverse effects. Sampling, for Vγ9/Vδ2 studies, were performed before the start of therapy and at the time of the second infusion.

All values are expressed as mean ± SD. We performed analysis of significance in Prism (GraphPad, La Jolla, CA, USA) by the two-tailed t test analysis and by two-way ANOVA.

Cumulative results are shown in Table 1. The EF of Vγ9/Vδ2 T lymphocytes from active BD patients was significantly higher (295 ± 44; P <0.0001) than those obtained from inactive patients (36 ± 8) and from healthy controls (30 ± 8). Vγ9/Vδ2 TNF-RII was 10,422 ± 1,694 in active patients, 4,087 ± 1,671 in inactive patients and 4,512 ± 1,436 in healthy controls. The intracytoplasmic content of IFN-γ was 40.4 ± 8.2% in Vγ9/Vδ2 standard cultures from active patients, and 18 ± 7% and 11 ± 4% in inactive patients and healthy subjects respectively. Intracytoplasmic perforin content was 27,581 ± 4,611 in active patients and, 5,247 ± 1,230 and 4,980 ± 1,110 in inactive patients and controls respectively.

GrA levels were (818 ± 97 mOD units) significantly higher in the supernatants obtained from cultures of active patients than those obtained from inactive patients (589 ± 21) and controls (550 ± 30). Staining of PBMC cultures with antibodies to Vδ2-TCR, CD45RO and CD27 identified four subsets of Vγ9/Vδ2 T cells (Table 2): 1) a subset with a naive CD45RO^- CD27⁺ phenotype, 2) a subset with a memory CD45RO⁺ CD27⁺ phenotype; 3) a subset with a effector CD45RO⁺ CD27^- phenotype and 4) a subset with a terminally differentiated cytotoxic CD45RO^- CD27^- phenotype. In active patients we observed that the phenotype of Vγ9/Vδ2 T cells is mainly composed of CD45RO⁺CD27^- effector cells (57 ± 13%) and that there is, also, evidence of CD45RO^-CD27^- terminally differentiated cytotoxic cells (20 ± 6%). In contrast, in Vγ9/Vδ2 cultures of inactive patients as well as of normal controls, we found mainly the phenotypes CD45RO⁺CD27⁺ (47 ± 9 and 49 ± 9%) and CD45RO^- CD27⁺ (24 ± 7 and 20 ± 8%) that are characteristics of memory and naive cells, respectively.

The addition of Infliximab to the cultures from active patients determined a significant time and dose dependent inhibition of cell expansion (Figure 1a) and a significant decrease of TNF-RII expression (Figure 1b). In particular, the mean EF was 295 ± 43.4 in untreated cultures and 50 ± 23 in culture with the highest infliximab concentration (P < 0.001). TNFRII MESF was 10,500 ± 1,707 in untreated culture and 4,576 ± 845 (Infliximab 10 μg/ml for 3 days; P < 0.05), 3,466 ± 394 (Infliximab 50 μg/ml for 3 days; P < 0.01) and 833 ± 165 (Infliximab 100 μg/ml for 10 days, P < 0.001) respectively after exposure to infliximab. IFN-γ in untreated cultures was 40.4 ± 8.25%. In the presence of Infliximab we found values of 10.5 ± 3.51% (Infliximab 10 μg/ml for 3 days; P < 0.01), 7.2 ± 1.8% (Infliximab 50 μg/ml for 3 days; P < 0.001) and 5.5 ± 1.7% (Infliximab 100 μg/ml for 10 days; P < 0.001) (Figure 1c).

A significant reduction of cell perforin content (7,040 ± 985 vs 27,573 ± 4,590, P < 0.001) was also observed after infliximab (50 μg/ml for 10 days) (not shown) exposure together with a decrease of granzyme release in the supernatants (Figure 1d); in standard cultures the GrA levels were 818 ± 91.6 mOD, in the presence of Infliximab the levels were 538 ± 72 (10 μg/ml for 3 days; P < 0.05), 518 ± 69 mOD (50 μg/ml for 3 days; P < 0.05) and 240 ± 25 mOD (100 μg/ml for 10 days; P < 0.01). The Vγ9/Vδ2 cells from active patients cultured in presence of Infliximab (50 μg/ml) shown mainly the phenotype of memory (75 ± 15%) and naive (21 ± 3%) cells rather than those of effector (1.76 ± 1.28%) and cytotoxic (0.3 ± 0.21%) cells (Figure 2).

Five patients with active disease were infused with 5 mg/Kg of Infliximab, and Vγ9/Vδ2 studies were performed before and after therapy (Table 3). A significant EF difference was found in the Vγ9/Vδ2 cultures from patients before (256 ± 90) and after (80 ± 44; P < 0.01) Infliximab therapy. Serum GrA levels were 715 ± 174 mOD and 207 ± 72 mOD (P < 0.001) respectively. After therapy, the cells showed the phenotype of memory (61 ± 11%) and naive (21 ± 3%) cells.